You can use csvread
as the data is numeric only. And access the lat/lon by standard column indexing.
M = csvread(filename);
lat = M(:,5);
lon = M(:,6);
Then select the indexes corresponding to your criteria and extract the data
% a N x 4 matrix, N lines for the number of groups, and 4 columns for min(lat) max(lat) min(lon) max(lon)
lat_lon_values = [10 20 50 80;
20 30 50 80;
50 70 10 60];
for ind_group = 1:size(lat_lon_values, 1)
a = lat_lon_values(ind_group, 1); %min(lat)
b = lat_lon_values(ind_group, 2); %max(lat)
c = lat_lon_values(ind_group, 3); %min(lon)
d = lat_lon_values(ind_group, 4); %max(lon)
%select the indexes based on lat and lon values
ind_line = M(5,:)>= a & M(5,:)<= b & M(6,:)>= c & M(6,:)<= d;
%your selected data
M(ind_line, :)
end